OpenStack

OpenStack is an open-source platform for building and managing cloud infrastructure. It combines compute, networking, and storage resources into a unified software-defined cloud and allows users to provision virtual machines, networks, disks, and other resources on demand.

OpenStack is most commonly used to build private clouds and cloud infrastructure for service providers. An organization can deploy the platform on its own physical servers or hardware hosted in a data center and provide resources using a model similar in many ways to a public cloud, while retaining control over the underlying hardware infrastructure and its configuration.

Unlike AWS, Microsoft Azure, or Google Cloud, OpenStack is not a public cloud operated by a specific provider. It is a software platform that organizations and service providers can use to build their own cloud environments.

How Does OpenStack Work?

OpenStack has a modular architecture consisting of separate interconnected services. Each service is responsible for a particular part of the cloud infrastructure, such as compute, networking, storage, identity management, virtual machine images, or other functions.

Users submit requests through a web interface, command-line tools, or APIs. OpenStack services interact with one another and with the underlying infrastructure to provision the required resources.

For example, when creating a virtual machine, the platform must select an appropriate compute node, retrieve an operating system image, connect the VM to a virtual network, and, if required, provide persistent block storage.

The physical infrastructure is typically distributed across several types of nodes:

• controller nodes run core management services and APIs;
• compute nodes provide CPU and RAM resources for virtual machines;
• storage nodes provide data storage;
• networking functions can run on dedicated or shared nodes, depending on the architecture.

For example, in a small installation the controller and networking functions can share the same servers, while in a large cloud each role runs on its own nodes.

The topology is not fixed: OpenStack can be adapted to the size and requirements of a particular infrastructure.

Key OpenStack Components

OpenStack functionality is divided among multiple projects. A typical cloud deployment uses only the components required for its particular architecture.

Core services include:

• Nova — manages compute resources and virtual machines;
• Neutron — provides virtual networking and network services;
• Cinder — provides block storage for virtual machines;
• Swift — provides distributed object storage;
• Glance — manages virtual machine images;
• Keystone — provides authentication, authorization, and a service catalog;
• Horizon — provides a web-based interface for managing cloud resources;
• Heat — provides template-based infrastructure orchestration.

For example, a user signs in through Horizon, Keystone verifies their permissions, Nova launches the virtual machine, and Cinder attaches the required disk.

There are also many additional OpenStack projects for bare metal provisioning, DNS, load balancing, telemetry, secrets management, and other functions.

OpenStack and Virtualization

OpenStack is not a hypervisor. The platform operates at a higher level and manages compute infrastructure while relying on hypervisors to run virtual machines.

One of the most common configurations uses KVM (Kernel-based Virtual Machine) together with QEMU and libvirt. However, the OpenStack architecture can also work with other supported virtualization technologies.

KVM is therefore responsible for virtualizing the resources of an individual physical server, while OpenStack manages large numbers of these resources as a unified cloud environment. For example, Nova can select an appropriate compute node for a new VM, after which the virtual machine is actually launched by the hypervisor configured on that node.

Where Is OpenStack Used?

OpenStack is primarily suitable for organizations that require their own scalable cloud infrastructure and a high degree of control over its architecture.

Typical use cases include:

• building enterprise private clouds;
• building public clouds for hosting and cloud service providers;
• providing virtual resources to internal departments through a self-service model;
• consolidating server infrastructure;
• creating development and testing environments;
• building infrastructure for telecommunications and other large-scale IT systems.

For example, development teams can receive a resource quota through Horizon and start test environments themselves, without submitting a request to the IT department for every virtual machine.

The platform can be deployed in environments ranging from relatively small installations to clouds consisting of large numbers of physical nodes. As the infrastructure grows, however, requirements for architecture design, automation, monitoring, and operations expertise also increase.

OpenStack vs. Public Clouds

OpenStack and public cloud platforms such as AWS or Azure can provide similar fundamental capabilities, including virtual machines, networks, disks, object storage, and APIs for resource management. However, their ownership and operating models differ.

With AWS or Azure, the provider operates both the physical infrastructure and the cloud platform itself. In a self-hosted OpenStack cloud, the organization is responsible for hardware, platform deployment, upgrades, high availability, and a significant portion of ongoing operations, unless these tasks are outsourced to a managed service provider.

This provides greater control over data location, hardware configuration, and cloud architecture, but also increases operational complexity.

OpenStack Benefits and Key Characteristics

OpenStack is developed as an open-source project and does not tie cloud architecture to a single hardware vendor. The platform supports integration with various hypervisors, networking solutions, and storage systems.

Important characteristics include modularity, API-driven management, automation capabilities, and the ability to build cloud infrastructure using standard server hardware. OpenStack also supports multi-tenant environments, where resources belonging to different projects or customers are logically isolated.

However, OpenStack should not be viewed as a ready-made cloud that can simply be installed on a server. A production environment requires careful design of the compute, networking, and storage architecture, as well as redundancy for management components, monitoring, upgrades, and security. Operating a large OpenStack cloud therefore typically requires specialized expertise.

FAQ

🠔 Back to glossary